OpenAI launches a mini version of its latest AI model.
These small AI models are intended to be faster and cheaper than the complete version – and are due to this fact particularly suitable for easy, complex tasks. This should appeal to smaller developers who don’t necessarily have a variety of money for AI costs, but need to integrate AI into their website or app relatively easily.
In this case, OpenAI announced its latest flagship model, GPT-4o, back in May. The “o” stands for “omni,” a sign that the model should have the opportunity to grasp speech and video along with text.
And now there's GPT-4o mini, the stripped-down version. GPT-4o mini currently supports text and pictures, but OpenAI says it would add video and audio capabilities in the longer term. This recent mini model is alleged to be greater than 60% cheaper than GPT 3.5 Turbo, which it replaces as OpenAI's smallest model. It also performs higher on MMLU (an industry benchmark for logical reasoning) than competing small models.
Click play to learn more, then tell us what you think that within the comments!
